Juitmal of Orthoputd1,c Rrsrart li

1061M20 Raven Press. Ltd , New York

0 1992 Orthopaedic Research Society

Local Stimulation of Proteoglycan Synthesis in Articular Cartilage Explants by Dynamic Compression In Vitro Jyrki J. Parkkinen, Mikko J. Lammi, Heikki J. Helminen, and Markku Tammi Department of Anatomy, University of Kiiopio, Kuopio. Finland

Summary: Cultured bovine articular cartilage was subjected to 50 ms, 0.5-1.0 MPa compressions repeated at intervals of 2-60 s for 1.5 h and simultaneously labeled with %O,. The compression was delivered with a 4-mm-diameter nonporous loading head on an 8-mm-diameter cartilage explant. This method created directly compressed (central) and uncompressed (border) areas within the tissue. Analysis of the whole explant under a 0.5 MPa load showed significantly increased "SO, incorporation by compression repeated at 2- and 4-s but not at 20- and 60-s intervals. When the incorporation was studied separately in the border and central areas, a statistically significant stimulation was noticed in the central area with a 4-s cycle, while the border area was stimulated with a 2-s cycle. Autoradiography of the central area showed that the stimulation with 0.5 MPa and a 4-s cycle occurred through the whole depth of the cartilage, while raising the pressure to 1 MPa or the frequency to 2 s reduced the stimulation, particularly in the superiicial cartilage. In the border area the stimulation with 0.5 MPa and a 2-s cycle was noted in the superficial zone only. The stimulation of proteoglycan synthesis is thus limited to certain loading frequencies and pressures and occurs in specific areas under and around the loaded site. Its rapid appearance suggests enhanced glycosylation or sulfation of core proteins or enhanced speed of posttranslational processing. Key Words: Autoradiography-In vitro loading-Proteoglycan biosynthesis.

drostatic, and stretching forces in an isolated system with controlled force values (3,7,8.9,13,16,21, 22,24,25,36,37,42-44,49,50). In general, the cited studies show that when the tissue is exposed to static, deformative stress, proteoglycan synthesis decreases, while intermittent loading stimulates it, a5 indicated by [35S]sulfateincorporation. The aim of this study was to assess in more detail proteoglycan synthesis stimulation during dynamic loading. We used a recently constructed apparatus in which the magnitude and frequency of the mechanical force is freely selccted and accurately controlled (37). The idea that local stress values influence the synthesis of proteoglycans by chondrocytes was probed by separate analyses of 35S0, incorporation in the tissue directly under the loading head and in the tissue around the loading head. Furthermore, quantitative autoradiography on tis-

Data from numerous experiments on various animal models suggest that joint loading exerts a strong influence on the development, structure, and metabolism of articular cartilage (48). Studies on various sites of articular surfaces suggest that local contact pressures during joint loading control the synthesis and degradation of the extracellular matrix by resident chondrocytes (19,46,47). However, in vivo experiments alone cannot confirm this idea or characterize the mechanisms that play a part at the cellular level. In vitro methods ofTer a good way to study the response of articular cartilage to compressive, hyReceived February 1 , 1991: accepted April 20, 1992. Address correspondence and reprint requests to Dr. J. J. Parkkinen at Department of' Anatomy. University of Kuopio, P.O.B. 1627, 70211 Kuopio. Finland.

610

CARTILAGE COMPRESSION IN VITRO

sue sections was used to measure the incorporation at different depths of the cartilage. MATERIALS AND METHODS Tissue Culture Bovine knees from 1-2-year-old animals were obtained from the local abattoir (Lihapolar, Kuopio, Finland). The joint was opened under sterile conditions, and two cartilage plugs (8 mm in diameter) were punched from the lateral side of the patellar surface of femur. Subchondral bone and calcified cartilage were separated with a scalpel, and the plugs of uncalcified cartilage were immediately attached to a plastic culture dish with Histoacryl tissue adhesive (B. Braun Melsungen, Melsungen, Germany). One of the explants was exposed to compression during in vitro loading, while the other one served as a control sample (Fig. 1A). The cartilage explants were cultured at 37°C in Eagle’s MEM with Earle’s salt (Flow Lab Ltd., Irvine,

Loading head

....

B

. . . . . . . . ,

4mm

,

.

Central area

611

Scotland) supplemented with antibiotics (1 00 U/ml penicillin and 100 pdml streptomycin; Flow Lab), 70 pg/ml ascorbate (Sigma, St. Louis, MO, U.S.A.), and 3 mM glutamine (Flow Lab). The explants were generally allowed to stabilize in culture before the experiments. Proteoglycan synthesis was examined to uncover any alterations in the synthesis rate during this period. Articular cartilage plugs were attached to culture dishes as described above and precultured for 0, 6, 24, or 60 h (n = 11 in each group) at 37°C before labeling with 50 pCi/ml of [3sS]sulfate (carrier free; Amersham Intl., Little Chalfont, England) for 1.5 h. The plugs were digested overnight with papain (Sigma) at 60°C. The incorporated 35S radioactivity was isolated by PD-10 columns (Sephadex G-25; Pharmacia, Sweden) and measured by a liquid scintillation counter (LKB, Bromma, Sweden). The linearity of sulfate incorporation for short labeling periods was studied in a 48 h preculture and 0.25-8.0 h labeling of the explants with 50 pCi/ml of [3sS]~ulfate. Diffusion rate into the cartilage plugs was determined by first killing the chondrocytes in two cycles of freezing and thawing. Two plugs were attached to the petri dishes as in other loading experiments, one for loading and the other as a control. The samples were subjected to a 1 .O-MPa load with a 50-ms peak load in a 2-s total cycle length for 5 , 15,30,60, or 90 min (10 pairs of explants in each group). Immediately before loading, 10 pCi/ml [3sS]sulfatewas applied to the dish. The loaded and control explants were wiped and separately rinsed (4 X 10 min) in medium and digested with papain, and the activities in the rinsing and digestion media were counted by liquid scintillation.

In Vitro Loading Apparatus

..

. .

.

. . .

...

,

Explant

FIG. 1. The loading system of articular cartilage explants. A: A pair of uncalcified cartilage plugs was taken from the patellar surface of bovine femur and attached to a petri dish for tissue culture. One of the explants was compressed with a microprocessor-controlled loading head, while the other served as control. B: After loading, both explants were detached, and the central and border areas were separated with a punch for scintillation counting and microscopic autoradiography.

In vitro loading was produced by an apparatus developed for cyclic compression of relatively large cartilage explants (diameter 5-10 mm). The device contains a stepping motor that moves the 4-mmdiameter, nonporous, stainless steel loading head against a cartilage explant, glued onto a petri dish, and load cell under the dish to measure the compressive pressure (Fig. 2 . ) . Because the loading head diameter was smaller than the explant, in most experiments the explant was divided into a directly compressed central area and an uncompressed border area. A microprocessor controls the stepping motor with the help of a feedback signal from the load cell. The stress can be selected in a range from

J Orthop Res, Vol. 10, No. 5 , 1992

612

J . J . PARKKINEN ET AL. center (contact) and border (noncontact) areas of the explants were separated with a punch (diameter of 5 mm) before analysis (Fig. IB) (n = 12 pairs in each group). Both areas were divided with a scalpel for incorporation and autoradiographic analysis. The samples for autoradiography ( 1 mm wide) were taken in the middle of the border and central areas and rinsed in cold phosphate-buffered saline (PBS), supplemented with 1 mM sodium sulfate for 3 x 10 min. The samples for the 35S incorporation assays were rinsed in PBS containing protease inhibitors. The wet weight of the samples taken for "S incorporation were recorded, and the tissues were digested overnight with papain at 60°C. The incorporated 35S radioactivity was isolated by PD-10 columns and measured by liquid scintillation. Sample Preparation for Autoradiography

FIG. 2. The loading apparatus. The shaft (Sh) of the linear moving unit (LMU) pierces the supporting plate on the top of the incubation chamber. The interchangable, plain-ended loaded head (LH) is attached to the shaft. The petri dish is fixed by screws to the socket (S) of the load cell (L). Continuous gas flow is delivered into the chamber.

25 kPa to 12.5 MPa, with an adjustable loadirest ratio. The total cycle length can be varied from 1 s to 59 s, and the peak loading phase can be a 1 5 ms. For further details of the apparatus, see Parkkinen et al. (37).

The samples were fixed in 2% glutaraldehyde in 0.1 M Na-cacodylate buffer (pH 7.4) for 4 h, postfixed in 1% osmiumtetroxide, dehydrated in ascending series of ethanol, and embedded in Epon (Ladd Research Industries Inc., Burlington, VT, U.S.A.). One-micrometer-thick (semithin) sections were cut with an ultramicrotome (Ultrotome I11 8800; LKB) perpendicular to the articular surface. The thickness of the sections was controlled by the Watson interference objective (Parry Instruments Ltd., Harpenden, England) (40). The sections were dipped in Kodak NTB-2 (Eastman-Kodak, Roches-

O.7

In Vitro Loading Experiments To study the effects of loading frequency on sulfate incorporation, the cartilage explants, after 2 days preculture, were subjected to 0.5-MPa cyclic compression for 1.5 h and simultaneously labeled with SO pCi/ml of [35S]sulfate(n = 12 pairs in each group). The duration of maximum stress was always 50 ms in 2 s, 4 s, 20 s, or 60 s of total cycle length (Fig. 3). In experiments aimed at studying the local variation of the cartilage response to loading of 0.5- and 1.0-MPa stress with 2-s and 4-s cycle lengths, the

J Orthop Res, Vol. 10. No. 5 , 1992

0 0

0.5

1

1.5

2

2.5

3

Time (s) FIG. 3. The loading profiles used for the explants. The curve shows the pressure recorded as a function of time in an experiment where the pressure time was set at 50 ms, the load was 0.5 MPa, and the cycle length was 2 s. The microprocessor of the device automatically controlled the loading head through feedback information from a load cell under the culture dish. For further details see (37).

613

CARTILAGE COMPRESSION IN VlTRO

ter, NY, U. S .A .) autoradiography emulsion diluted 1:1 with deionized water. After 2 weeks’ exposure at 4°C the autoradiographs were developed in Ilford PQ universal developer and fixed with Hypam rapid fixer (Ilford Ltd., Mobberley, England) to 4.5 min in each solution. The sections were lightly stained with 0.2% ruthenium red in 0.1 M cacodylate buffer, pH 7.4, and 1% phenylenediamine in 0.1 M phosphate buffer, pH 7.4, for 5 min in each (38) and mounted with D.P.X. “8711” (Difco, East Molesey, England). Autoradiographic Analysis The procedure for the quantitation of the autoradiographic grains has been described in detail previously (38). The method is based on separation of grains from the underlying structures by gray level thresholding and determination of the percentage of the total area occupied by the grains. The mean grain size was used as a reference for calculation of grain number per unit area of tissue in various zones of articular cartilage. Microscopic sections cut perpendicular to the cartilage surface were analyzed in consecutive fields (depth -90 pm) starting from the cartilage surface and ending at the deep zone. The pictures were transferred from a Vanox AH-2 light microscope (Olympus, Tokyo, Japan), equipped with a 40X objective, through a Tk-870E CCDvideo camera (JVC, Japan) to an IBAS image analyzer (Kontron, Munich, Germany), and the relative area of grains in the digitized fields were calculated by the image analyzer (38). Strain Approximation During loading experiments, the deformation of the cartilage plug was not measured. However. the strain was theoretically approximated by using the mathematical formulation of Hayes et al. (11) for the relationship between the applied load and displacement of articular cartilage. During short-term loading, articular cartilage behavior can be satisfactorily modeled to be equal to that of an incompressible elastic material (14). For an elastic material bonded to a rigid base and compressed with a cylindrical and plain-ended loading head (11) we obtain w

=

(P(1

-

v)/(4 a

K p)

where w is deformation, P is load, u is Poisson’s ratio (u = 0.50 for incompressible isotropic elastic

material) (14), a is loading head radius, K is a theoretical scaling function of integral form, and p is shear modulus. The value of K depends on the areaaspect ratio ( d h ; h = cartilage thickness) and Poisson’s ratio (11,15). Typical values for the thickness and shear modulus of bovine patellar groove cartilage were 1.5 mm and 5.1 MPa, respectively (41), which were used for approximation of the maximum strain during loading experiments. The strain imposed on the cartilage plug during loading experiments was checked with a cartilage testing device, a modification of the present loading apparatus (41). The conditions and protocol for testing was exactly the same as used during loading experiments. The position of the loading head was monitored by an optoelectronic displacement transducer connected to a microcomputer through 12-bit AID counter. For calculation of the strain, the thickness of cartilage was measured by a penetrating needle technique (12). Statistical Analysis Student’s paired t test was used for statistical analysis. In the figures, a single asterisk represents a p value < 0.05, and a double asterisk represents p < 0.01. RESULTS Characterization of the Culture Conditions The cartilage plug was separated from the subchondral bone near tidemark and immediately attached onto a petri dish by a small amount of tissue adhesive to prevent lateral sliding of the explant during loading. As judged from light microscopic observations of histological sections, the adhesive did not penetrate into the tissue. After hardening, the stiffness of the adhesive was found to be much higher than that of articular cartilage. Because of the stiffness and extremely thin layer of the adhesive, all deformation was considered to occur in the cartilage plug. To test the culture conditions and possible alterations in proteoglycan synthesis after cutting the explants, the 3sS04 incorporation rate was rnonitored during the first 2 days in culture. The synthesis rate of proteoglycans displayed no significant alteration during this time (Fig. 4). The kinetics of 35S04incorporation was determined in explants cultured for 2 days before labeling for 0.25-8 h. The

J (Irthop R F J , Vol. 10, N o . 5, 1992

614

J . J . PARKKINEN ET AL. 6000 1

-3

.-0 ) a,

3

a, 2000

4000

.

E" -- 1000

= a

E a

0

0

2oool/ 0

0

0

0

24

6

15

48

30

45

75

60

90

Time (min)

Time (h) FIG. 4. Radiosulfate incorporation rates of the explants during the first 2 days in culture. The explants were labeled for 1.5 h at the indicated times after preparation of the tissue. The bars show SEM (n = 11)

incorporation rate was also linear with short labeling periods (Fig. 5). The diffusion rate of 3's04in loaded and control tissues was compared exactly as in the metabolic labelings with living explants by incubating killed cartilage. The amount of tissue 35S04reached 7080% of the equilibrium level within 5 min (Fig. 6). The rapid diffusion agreed with the linear incorporation found even after short labeling periods (Fig. 5). There were no significant differences in the equilibrium rate of 35S04between the loaded and the control explants. Loading and Proteoglycan Synthesis

In preliminary studies the area ratio of the loading head and the cartilage explant was varied in a series of experiments using 0.5-MPa pressure and a 2-s cycle. A higher average stimulation in sulfate incor-

FIG. 6. Diffusion rate of radioactive sulfate into the cartilage explants. The explants were frozen and thawed twice and then subjected to 1-MPa load at 2-s intervals for 5-90 min. The radiolabel was added at the beginning of the loading period. The loaded and control explants were wiped and separately rinsed (4 x 10 min) in medium and digested with papain, and the activities in the rinsing and digestion media were counted by liquid scintillation. The curves show the sum of the counts (mean 2 SEM, n = 10) in the rinsing and digestion media.

poration was observed with larger explants (diameter 8-10 mm) as compared with smaller explants (diameter 4-5 mm) when the loading head diameter was kept constant (diameter 4 mm). Therefore, explants of 8 mm diameter were used in subsequent experiments. First, the influence of the loading frequency on sulfate incorporation in whole explants was examined with 50-ms 0.5-MPa peak loads repeated at 2-s, 4-s,20-s, and 60-s intervals. As shown in Fig. 7 , the rapid cycles (2- and 4-s intervals) were stimulatory (n = 12 in both groups, p < 0.05), while loading at

*

c.

r U

Control Loaded

0

//

120001

0

0

2

4

6

8

Time (h) FIG. 5. The linearity of sulfate incorporation as a function of time. The cartilage explants were precultured for 2 days before labeling for 0.25-8 h.

J Orthop Res, Vol. 10, N o . 5 , 1992

6 100 6

v

H

a

O

n

U

2s

4s 20 s Cycle length

60 s

FIG. 7. The influence of loading frequency on radiosulfate incorporation in proteoglycans. A 0.5-MPa load of 50 ms in peak load time was applied to the cartilage for 1.5 h at 2-s, 4-s, 205,and 60-s intervals. The isotope (50 pCi/ml) was added at the beginning of the loading. The explants were proteolytically digested, and macromolecular radioactivity was isolated by gel filtration. The results are expressed as percentage of controls and represent the mean and standard error of 12 pairs of explants from separate animals. Student's paired t test was used for statistical analysis. *p < 0.05.

615

CARTILAGE COMPRESSION IN VlTRO

20- and 60-s intervals showed no statistically significant effect. Since the preliminary studies suggested that the width of the outer, noncontact area influenced sulfate incorporation of the whole explant, the incorporations were then determined both for the central tiswe under the loading head and for the border area of the explant, separated after loading with a punch (Fig. 1B). The effect of the stress magnitude (0.5 and 1.0 MPa) on the stimulation of sulfate incorporation was examined separately in central and border areas with 2-s and 4-s cycles (12 pairs in each). Under the loading head (i-e., the central area), by using a 4-s cycle length, incorporation was stimulated more with the lower (0.5 MPa) (p < 0.01) than with the higher (1 .O MPa) pressure. With the faster (2 s) cycle there was no statistically significant stimulation in the central area directly under the loading head (Fig. 8A,B). A significant stimulation of sulfate incorporation was also observed in the border area (p < 0.05) using a 2-s cycle and 0.5-MPa load (Fig. SD). The distribution of sulfate incorporation in the central and border areas was determined by automated counting of the autoradiographic grains in the different zones of the cartilage (Fig. 9). In general, the ratios of total grain numbers in sections from loaded and control explants (Fig. 10) corresponded to the cpm values in scintillation counting (Fig. 8). The grain density was highest in the intermediate zone and considerably lower both in the surface zone and in the deep zone (Fig. 10). The autoradiographic analysis of the central area showed higher grain numbers with 0.5-MPa than with 1 .Q-MPa pressures with both 4- and 2-s cycles (Fig. 10A,B). With a 0.5-MPa pressure and a 4-q cycle, the stimulation in the central tissue was noted in all zones (Fig. 1QA), but decreasing the cycle length to 2 s produced stimulation only in the intermediate zone (Fig. 10B). Using the higher pressure (1 .O MPa) only intermediate zone stimulation was noted with the 4-scycle, and with the 2-s cycle hardly any stimulation was left even in the intermediate zone (Fig. IOA,B). In the border area the increased 35S0,incorporation with the 2-s cycle and 0.5-MPa pressure (Fig. 8) was located in the superficial parts of cartilage, as determined by autoradiography (Fig. IOD). By other loading regimes there were only small differences in the border area between control and loaded groups.

300

Central area

]A

IB

4s

2s H Control

200

Loaded

c.

c

0 0

Y-

O

Border area

.a

2 300 2

a

0

200 100 fi

U '

0.5

1.0

0.5

1.0

MPa FIG. 8. Sulfate incorporations in the central and border areas of the explants (see Fig. l ) , loaded with 4-s (A,C) or 2-s (B,D) cycle lengths using 0.5-MPa or 1.O-MPa pressure. Loading time and labeling were the same as in Fig. 7. The results represent the mean f SEM of 12 pairs of explants at each point. Student's paired t test was used for statistical analysis. *p < 0.05. **p < 0.01,

Using theoretical modeling the maximum strains were calculated to be 1 % of cartilage thickness with a 0.5-MPa load and 2% with a 1-MPa load. These theoretical strains were comparable to those found by experimental measurements.

DISCUSSION In this study a recently designed loading apparatus (37) was applied to investigate the influences of different magnitudes and frequencies of loading on the incorporation of 3 5 S 0 , in cultured articular cartilage. Unlike many previous in vitro loading systems, a large explant (diameter of 8 mm) was compressed with a relatively smaller loading head (diameter of 4 mm), which created directly compressed and uncompressed sites within the same tissue. By separate analyses of the directly compressed (central area) and noncontact (border area) sites, and by a novel quantitative analysis system of

J Orthop Res, Vol. 10, No. 5 , 1992

616

J . J . PARKKINEN ET A L .

FIG. 9. An autoradiograph from the central area of a loaded sample (4 s, 1 MPa). The bar represents the depth of the superficial zone. The corners show the uppermost measuring field, which corresponded to the superficial zone. Original magnification x450.

autoradiographic grains in microscopic tissue sections, the topographical distribution of net proteoglycan synthesis under and around the loading head could be determined. The stresses acting in the bovine knee joint in vivo have not been investigated, but it has been calculated that the static compressive stress in the cow knee joint is 120 psi (0.8 MPa). The stresses on the cartilage surface in species of widely different sizes are relatively constant (45). The pressures

3 Orthop Res, Vol. 10, No. 5 , 1992

used in our study (0.5 MPa, 1 .0 MPa) represent values at the lower end of the physiological pressures found in the human knee joint during walking (0.86.3 MPa) (4). The diffusion of radiosulfate into the explant was found to be rapid; 70-80% of the equilibrium concentration was achieved in 5 min (Fig. 6). The result is in agreement with the findings of Maroudas and Evans, who noted that in 1.5-mm-thick human articular cartilage 90% of the equilibrium concentra-

CARTILAGE COMPRESSION IN VITRO 0.5 MPa

617

0.5 MPa

1.0 MPa

1.0 MPa 2sl

20 10 0 L

.m

0

10

0

10

20

40

30

C

30

4s

10 0

0

10

0

10

20

Layers from surface FIG. 10. Influence of loading on autoradiographic grain numbers in different cartilage depth layers. Loading time, cycles, and loads were as in Fig. 8 . A,B: the central area. C,D: the border area. In each control explant the total grains in all layers were summed, and the proportion (percentage) of each layer was calculated. The figure shows the mean and SEM (n = 12) of these control distributions (-). In each loaded (-B) explant the total grain number was first calculated as a percentage of the corresponding control explant, and this percentage was distributed to the layers.

tion was reached within 20 min (28). There was no significant difference in the equilibration curve between loaded and control explants. Therefore, we concluded that a stimulation of 35S04incorporation by repeated loading is not due to a more rapid arrival of the isotope within the interior of the explant, neither does it seem probable that it is due to enhanced supply of small nutrients for the chondrocytes (36). We found no major alterations in 35S04incorporation within 2 days aftcr starting the culture, even though no serum or insulin-like growth factor (IGF) was used (10,31). Earlier studies have suggested that serum and IGF enhance sulfate incorporation and prevent the depletion of proteoglycans in cartilage explant cultures (10,31). It is not clear how these in vitro conditions correspond to those in vivo as to the effect of serum factors. One explanation of the maintenance of a constant proteoglycan synthesis in our culture system may be the relatively large size of the explants, which might slow down the diffusion of IGF and other growth factors from the tissue, as compared with the smaller explants used previously (10). The strain, as calculated from the linear elastic model of articular cartilage and confirmed by experimental measurements (n = 3), varied between 1 and 2% of cartilage thickness. Because most of the strain affected the very superficial tissue, its deformation is important. Since the actual loading time was short, the general fluid flow remained small.

The plain-ended loading head created complex stresses in the cartilage. The stress distribution in the transitional area between central and border areas is particularly difficult to estimate. However, it is typical of concentrated loads, as is the case with our loading experiments (area-aspect ratio = 1.33), that high tensile stresses arise at the articular cartilage surface near the periphery of the loaded area. On the other hand, with a small area-aspect ratio the load is spread into the cartilage plug, and the perpendicular stress remains rather low in the deeper parts of cartilage under the loading head (1). Since the cartilage plug is immobilized to the rigid base, shear stresses and strains arise under the edge of the loading head at the cartilage-base interface (1). In fact, the contact between loading head and articular surface is not totally frictionless; shear stresses are also obvious at the cartilage surface. Our data show that in the central area compression every 2 s is less stimulatory than compression every 4 s. The reduction of incorporation by the higher loading frequency was noted particularly in the superficial part of cartilage. Most of the deformation takes place in the superficial zone of articular cartilage compared with deeper layers (6,34), possibly because of the lower proteoglycan concentration in the surface (18). By deformation of the superficial parts of cartilage, it can be anticipated that its fixed charge density increases and pH decreases, both being factors that inhibit sulfate incorporation (8). Superficial chondrocytes are also de-

J Orlhop Res, Vol. 10, N o . 5 , 1992

618

J. J. PARKKINEN ET A L .

formed under direct compression, while chondrocytes in the deeper parts largely avoid this problem. It may be that with the higher loading frequency there is not enough time for rehydration of the compressed superficial cartilage to maintain optimum conditions for proteoglycan synthesis. It is interesting that higher frequcncies and loads were generally required to significantly stimulate incorporation in the border area as compared with the central area (Fig. 8) and that the stimulation in the border area occurred particularly in the superficial zone (Fig. 10). It is obvious that superficial tensile strains, opposed by stretching of the tangential collagen fibers. also modulate proteoglycan metabolism in the superficial parts of the border area cartilage. The most pronounced stimulation in 3sS04incorporation by serum occurs in the superficial zone (20,27,30), which is only faintly labeled in normal cartilage in vivo (29,32). After 7 days dynamic loading (0.2 MPa, 0.3 Hz) of the cartilage of calf sesamoid bone, proteoglycan synthesis in the superficial zone increased, as compared with day 0 and with unloaded control in day 7 (22). However, incubation in serum-enriched medium without loading also increased proteoglycan synthesis in the superficial zone. Penetration of serum IGF into cartilage is probably limited because of the large size of the IGF protein complex (30). While the pumping action of loading has little effect on small molecules, it enhances the diffusive movement of macromolecules (35). The superkicial chondrocytes may thus have a better supply of the growth factors that stimulate proteoglycan synthesis when serum is present. The zones of articular cartilage also differ in their response to joint loading in vivo (19,39). A considerable increase in proteoglycan content was observed in the intermediate zone of canine articular cartilage after moderate running exercise, while more strenuous running exercise voided the stimulation in the intermediate zone and caused proteoglycan depletion in the superficial Lone (17). A similar response was seen in our in vitro experiments, where a surplus of pressure (from 0.5 to 1.0 MPa) and loading frequency (load interval from 4 to 2 s) reduced the stimulation of sulfate incorporation, the superficial parts of cartilage being most sensitive to inhibition (Fig. 10). In these experiments the loading time was 1.5 h, a relatively short observation period compared with those reported previously. However, it is known

J Orthog Res, Vol. 10, No. 5 , 1992

that chondrocytes respond rapidly to physical force; the content of cyclic AMP rose to a maximum in 3 min after application of tensile stress (49). Pulse chase studies have shown that the average synthesis of the large proteoglycan takes 70-90 min (33), the last 10-15 min of which is spent on glycosylation, sulfation, and secretion into the extracellular space (26). Therefore, it is interesting that the stimulation of 35S04incorporation that we found may not involve de novo synthesis of core proteins but mainly increased the number, length, or sulfation of the glycosaminoglycan chains or enhanced the speed of processing of existing core proteins. It does not seem likely that the effect of loading is restricted to proteoglycan synthesis, since in other studies the incorporation of proline is stimulated to about the same degree as that of sulfate in shortterm loading experiments (9,43). While the experimental systems and pressures vary between different studies, there is always an upper limit for stimulatory pressures (9,43). A pure hydrostatic load (9) seems to remain stimulatory for higher pressure rates than a dynamic, mechanical one (43). Dynamic loads, as used in this and previous studies (43), create a temporary hydrostatic pressure within the cartilage but have additional influences related to the deformation of the tissue. These include increased proteoglycan concentration, decreased pH (7), fluid flow, streaming currents, and potentials (5,23) and even injury of the chondrocytes and collagenous network ( 2 ) . It may be that one or more of these additional components of dynamic, deformative loading inhibit the stimulation that otherwise would be obtained with the hydrostatic pressure. However, to substantiate the idea of hydrostatic pressure as a stimulatory factor in proteoglycan synthesis, comparison of the dynamic and purely hydrostatic loading systems with identical tissue, culture, and analysis methods is necessary. Constructing a device that delivers cyclic hydrostatic pressures is in progress in our laboratory. Acknowledgment: This work was supported by grants from the University of Kuopio, the North Savo Fund of the Finnish Cultural Foundation, the Research and Science Foundation of Farmos, the Paulo Foundation, the Academy of Finland, and the Finnish Research Council for Physical Education and Sports, Ministry of Education. The authors thank Dr. Jukka Jurvelin, Ph.D., and Dr. Tuomas Rasanen, M.Sc., for their critical comments concerning the manuscript. The authors acknowledge Mrs. Eija Rahunen, Mrs. Elma Sorsa, Mrs. Eija

CARTILAGE COMPRESSION I N VITRO Voutilainen, and Ms. Eija Antikainen for skillful technical

assistance.

20.

REFERENCES 1. Askew MJ. Mow VC: The biomechanical fiinction of the collagen fibril ultrastructure of articular cartilage. JBiomech Eng 100:105-115, 1978 2. Broom ND. Myers DB: A study of the structural response of wet hyaline cartilage to various loading situations. Connect Tissue Res 7:227-237, 1980 3. Copray JCVM, Jansen HWB, Duterloo HS: Effects of compressive forces on proliferation and matrix synthesis in mandibular condylar cartilage of the rat in vitro. Arch Oral Biol 30:29%304. 1985 4. Finlay JB, Repo RU: Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Trans Biomed Eng 25:3439, 1978 5 . Frank EH, tirodzinsky ,4J:Cartilage electromechanics. 1. Electrokinetic transduction and effects of electrolyte pH and ionic strength. J Biomech 20:615427. 1987 6. Gore DM. Higginson GR. Minns RJ: Compliance of articular cartilage and its variation through the thickness. Phys Med Biol 28:233-247, 1983 7. Gray ML, Pizzanelli AM. Grodzinsky AJ, Lee RC: Mechanical and physicochemical determinants of the chondrocyte biosynthetic response. J Urthop Res 6:777-792, 1988 8. Gray ML, Pizzanelli AM, Lee RC, Grodrinsky AJ, Swann U.4: Kinetics of the chondrocyte biosynthetic response to compressive load and release. Riochem f3iophy.v Act0 991 : 415425, 1989 9. Hall AC, Urban JPG, Gehl KA: The effects of hydrostatic pressure on matrix synthesis in articular cartilage. J Orfhop Res 9:l-10, 1991 10. Hascall VC, Handley CJ. McQuillan DJ. Hascall GK. Robinson HC, Lowther DA: The effect on serum of biosynthesis of proteoglycans by bovine articular cartilage in culture. Arch Biochrm Biophys 124:20&2?3, 1983 11. Hayes WC, Keer LM, Herrman G. Mockros LF: A mathematical analysis for indentation tests of articular cartilage. J Biowzech 5541-551, 1972 12. Hoch DH. Grodzinsky AJ, Kooh TJ. Albert ML, Eyre DR: Early changes in material properties of rabbit articular cartilage after meniscectomy. J Orthop Res 1:4-12. 1983 13. Jones IL, Klarnfeldt A , Sandstrom T: The effect of continuous mechanical pressure upon the turnover of articular cartilage proteoglycans in vitro. Clin Orthop 165:283-289. 1982 14. Jurvelin J, Kiviranta I, Arokoski J , Tammi M, Helminen HJ: Indentation study of the biomechanical properties of articular cartilage in the canine knee. Eng Mrd 16:15-22, 1987 15. Jurvelin J, Kiviranta I , Saiamancn A-M, Tammi M, Helminen HJ: lndentation stiffness of young canine knee articular cartilage-influence of strenuous joint loading. J Biomeclz 23: 123%1246. 1990 16. Kampen van GPJ, Veldhuijzen JI’. Kuijer R. Stadt van de RJ, Schipper CA: Cartilage response to mechanical force in high-density chondrocyte cultures. Arthritis Rheum 28:419424, 1985 17. Kiviranta I: Joint loading influences on the articular cartilage of young dogs: Quantitative histochemical studies on matrix carbohydrates [Thesis]. Kuopio, Finland, Pub Univ Kuopio, Medicine, 1987 18. Kiviranta I, Jurvelin J , Tammi M, Saamanen A-M, Helminen HJ: Microspectrophotometric quantitation of glycosaminoglycans in articular cartilage sections stained with safranin 0. Hisfochemistn~82:24Y-255, 1985 19. Kiviranta I, Tammi M, Jurvelin J , Saamanen A-M, Helminen HJ: Moderate running exercise augments glycosamino-

21.

22.

23.

24. 25. 26.

21.

28.

29. 30.

31.

32. 33. 34.

35. 36. 37.

38.

619

glycan and thickness of articular cartilage in the knee joint of young beagle dogs. J Orthop Res 6:188-195, 1988 Korver GHV, van de Stadt RJ. van Kampen GPJ, van der Korst JK: Composition of proteoglycans synthesized in different layers of cultured anatomically intact articular cartilage. Matrix 10:394-401, 1990 Korver GHV, van de Stadt RJ, van Kampen GPJ, van der Korst JK: The effect of 1 week cyclic mechanical compression on cultured anatomically intact articular cartilage. 11. Effects of increasing loads [Abstract]. Proceedings of the THvlf’rhMeeting of FECTS. Bialystok, Poland, 86: 1990 Korver GHV, van de Stadt RJ. van Kampen GPJ, van der Korst JK: The effects of loading on the synthesis of proteoglycans in different layers of anatomically intact articular cartilage. Arthritis Rheum (in press) Lee RC. Frank EH. Grodzinsky AJ, Roylance DK: Oscillatory compressional behavior of articular cartilage and its associated electromechanical properties. J Biomech Eng 103:28&292. I98 1 Lee RC, Rich JB, Kelley KM. Weiman DS. Matthews MB: A comparison of in vitro cellular responses to mechanical and electrical stimulation. A m Surg 48567-574. 1982 Lippiello L, Kaye C, Neumata T, Mankin HJ: In vitro metabolic response of articular cartilage segments to low levels of hydrostatic pressure. Connect Tissue Res 13:99-107. 1985 Lohmander LS, Kimura JH: Biosynthesis of cartilage proteoglycan. In: Articular Cartilage Biochemistry, ed by KE Kuettner, R Schleyerbach, VC Hascall, New York, Raven Press. 1986, pp 93-1 1 I Maroudas A, Chriqui C, Weinberg P: The stimulation of proteoglycan synthesis in cartilage by serum: variation with depth and influence of cyclic compression. Trans Orthop Res Soc 13:72, 1988 Maroudas A, Evans H: Sulphate diffusion and incorporation inio human articular cartilage. Biochrrn Biophys Acta 338: 265-279, 1974 Maroudas A: Glycosarninoglycan turn-over in articular cartilagc. f ‘ h i h Trans Royal Soc London Series B 271:293313, 1975 Maroudas A, Schneiderman R, Weinberg C: Comparison between enhancing effects of serum and insulin on gag synthesis in different zones of cultured human articular cartilage. T r m s Orthop Kes Soc 15:315. 1990 McQuillan DJ, Handley CJ, Campell MA, Bolis S, Milway VE, Herington AC: Stimulation of proteoglycan biosynthesis by serum and insulin-like growth factor-I in cultured bovine articular cartilage. Biochenz J 240:423-430. 1986 Meachim G: The effect of scarification on articular cartilage in the rabbit. J R o n e Joint Surg [Br] 45:150-161, 1963 Mitchell L), Hardingham T: The effects of cycloheximide on the biosynthesis and secretion of proteoglycans by chondrocytes in culture. Biochern J 196521-529, 1981 O‘Connor P, Orford CR, Gardner DL: Differential response to compressive loads of zones of canine hyaline articular cartilage: Micromechanical. light and electron microscopic studies. Ann Rheum Dis 47:414420, 1988 O‘Hara BP, Urban JPG, Maroudas A: Influence of cyclic loading on the nutrition of articular cartilage. Ann Rheum llis 49536-539, 1990 Palmoski MJ, Brandt KD: Effects of static and cyclic compressive loading on articular cartilage plugs in vitro. Arthritis Rheum 2 7 5 7 5 4 8 I , 1984 Parkkincn JJ, Lammi MJ, Kajalainen S, Laakkonen J, Hyvarinen E, Tiihonen A. Helminen HJ, Tammi M: A mechanical apparatus with microprocessor controlled stress profile for cyclic compression of cultured articular cartilage. J Biomech 22:1285-1291, 1989 Parkkinen JJ, Paukkonen K , Pesonen E, Lammi MJ, Markkanen s, Helminen HJ, Tammi M: Quantitation of autordd-

620

39.

40. 41.

42.

43. 44.

J . J . PAR KKINEV ET AL.

iographic grains in different zones of articular cartilage with image analyzer. Histochetnistry 93:241-245, 1990 Paukkonen K , Helminen HJ: Chondrocyte ultrastructure in exercise and experimental osteoarthrosis: A stereological morphometric study of articular cartilage of young rabbits using transmission electron microscopy. Clin Orthop 224: 284-288, 1986 Robertson WM, Storey B, Griffiths BS: An interference technique for measuring the thickness of semi-thin and thick sections. J Microsc 133:121-124, 1983 Rasanen T , Arokoski J, Jurvelin J, Helminen HJ: Microcomputer-controlled cartilage testing device [Abstract]. In: Proceedings of the Seventh Meeting ofrhe European Society of Biomechanics, ed by A Odgaard, P Kjaersgaard, JO Sajbjerg, University of Aarhus, Aarhus, Denmark, p 13, 1990 Sah RL, Grodzinsky AJ, Plaas AHK, Sandy 513: Effects of tissue compression on the hyaluronate-binding properties of newly synthesized proteoglycans in cartilage explants. Biochem J 267:803-808, 1990 Sah RL, Kim Y, Doong JH, Grodzinsky AJ, Plaas AHK, Sandy JD: Biosynthetic response of cartilage explants to dynamic compression. J Orthop Res 7:61%636, 1989 Schneiderman R, Keret D, Maroudas A: Effects of mechanical and osmotic pressure on the rate of glycosaminoglycan synthesis in the human adult femoral head cartilage: An in vitro study. J Orthop Res 4:3931108, 1986

J Orthop Res, Vol. 10, No. 5 , 1992

45. Simon WH: Scale effects in animal joints. I. Articular cartilage thickness and compressive stress. Arthritis Rheum 13: 244-256, 1970 46. Saamanen A-M, Tammi M, Jurvelin J, Kiviranta I, Helminen HJ: Proteoglycan alterations following immobilization and remobilization in the articular cartilage of young canine knee (stifle) joint. J Orthup Res 8:863-873, 1990 47. Siiaiminen A-M. Tammi M, Kiviranta I, Jurvelin J, Helminen HJ: Levels of chondroitin-6-sulfate and nonaggregating proteoglycans at articular cartilage contact sites in the knees of young dogs subjected to moderate running exercise. Arthritis Rheum 32: 1202-1292, 1989 48. Tammi M. Paukkonen K, Kiviranta I, Jurvelin J, Saamanen AM. Helminen HJ: Joint loading-induced alterations in articular cartilage. In: Joint Louding: Eiology and Heulth of Articular Structures. ed by HJ Helminen, I Kiviranta, A-M Saarnanen. M Tammi, K Paukkonen, J Jurvelin, Bristol, John Wright and Sons, 1987, pp 64-88 49. Uchida A, Yamashita K, Hashimoto K, Shimomura Y: The effect of mechanical stress on cultured growth cartilage cells. Connect lissue Res 17:305-311, 1988 SO. de Witt MT, Handley CJ. Oakes BW, Lowther DA: In vitro response of chondrocytes to mechanical loading: The effect of short term mechanical tension. Connect Tissue Res 12: 97-109, 1984